fix code format

vllm/model_executor/layers/quantization/__init__.py

@@ -35,6 +35,8 @@ def get_quantization_config(quantization: str) -> Type[QuantizationConfig]:
         raise ValueError(f"Invalid quantization method: {quantization}")
 
     # lazy import to avoid triggering `torch.compile` too early
+    from vllm.model_executor.layers.quantization.quark.quark import QuarkConfig
+
     from .aqlm import AQLMConfig
     from .awq import AWQConfig
     from .awq_marlin import AWQMarlinConfig
@@ -56,7 +58,6 @@ def get_quantization_config(quantization: str) -> Type[QuantizationConfig]:
     from .neuron_quant import NeuronQuantConfig
     from .qqq import QQQConfig
     from .tpu_int8 import Int8TpuConfig
-    from vllm.model_executor.layers.quantization.quark.quark import QuarkConfig
 
     method_to_config: Dict[str, Type[QuantizationConfig]] = {
         "aqlm": AQLMConfig,

vllm/model_executor/layers/quantization/quark/quark.py

@@ -1,24 +1,23 @@
-import re
 import fnmatch
+import re
 from typing import Any, Dict, List, Optional, cast
 
 import torch
 
-from vllm.model_executor.layers.quantization.utils.quant_utils import (
-    FUSED_LAYER_NAME_MAPPING)
-from vllm.model_executor.layers.quantization.quark.utils import (
-    deep_compare, should_ignore_layer)
-
 from vllm.model_executor.layers.fused_moe import FusedMoE
-from vllm.model_executor.layers.quantization.quark.quark_moe import (  # noqa: E501
-    QuarkMoEMethod)
 from vllm.model_executor.layers.linear import (LinearBase, LinearMethodBase,
                                                UnquantizedLinearMethod)
-from vllm.model_executor.layers.quantization.quark.schemes import (
-    QuarkScheme, QuarkW8A8Fp8, QuarkW8A8Int8)
 from vllm.model_executor.layers.quantization.base_config import (  # noqa: E501
     QuantizationConfig, QuantizeMethodBase)
 from vllm.model_executor.layers.quantization.kv_cache import BaseKVCacheMethod
+from vllm.model_executor.layers.quantization.quark.quark_moe import (  # noqa: E501
+    QuarkMoEMethod)
+from vllm.model_executor.layers.quantization.quark.schemes import (
+    QuarkScheme, QuarkW8A8Fp8, QuarkW8A8Int8)
+from vllm.model_executor.layers.quantization.quark.utils import (
+    deep_compare, should_ignore_layer)
+from vllm.model_executor.layers.quantization.utils.quant_utils import (
+    FUSED_LAYER_NAME_MAPPING)
 from vllm.platforms import current_platform
 
 __all__ = ["QuarkLinearMethod"]
@@ -104,16 +103,17 @@ def from_config(cls, config: Dict[str, Any]) -> "QuarkConfig":
                     deep_compare(q_config, q_configs[0])
                     for q_config in q_configs):
                 raise ValueError(
-                    "The quantization method used for kv_cache should be the same, "
-                    "but the quantization method for the kv_cache layer in the "
-                    "config is different.")
+                    "The quantization method used for kv_cache should "
+                    "be the same, but the quantization method for the "
+                    "kv_cache layer in the config is different.")
             kv_cache_config = q_configs[0].get("output_tensors")
             if kv_cache_config is None:
                 raise ValueError(
                     "The kv_cache quantization configuration is empty.")
 
-            # Since we have already set kv_cache quantization configurations, we will remove
-            # the quantization configuration for the output_tensors corresponding to the kv_cache layer.
+            # Since we have already set kv_cache quantization configurations,
+            # we will remove the quantization configuration for the
+            # output_tensors corresponding to the kv_cache layer.
             for q_config in q_configs:
                 q_config["output_tensors"] = None
 
@@ -217,7 +217,7 @@ def _find_matched_config(self, layer_name: str,
         else:
             layer_quant_config = cast(
                 Dict[str, Any], self.quant_config.get("layer_quant_config"))
-            for name_pattern in layer_quant_config.keys():
+            for name_pattern in layer_quant_config:
                 if fnmatch.fnmatch(layer_name, name_pattern):
                     return layer_quant_config[name_pattern]
 

vllm/model_executor/layers/quantization/quark/quark_moe.py

@@ -1,9 +1,9 @@
-from typing import Callable, Optional, Dict, Any
+from typing import Any, Callable, Dict, Optional
 
 import torch
+
 import vllm.model_executor.layers.fused_moe  # noqa
 from vllm import _custom_ops as ops
-
 from vllm.model_executor.layers.fused_moe import (FusedMoE, FusedMoEMethodBase,
                                                   FusedMoeWeightScaleSupported)
 from vllm.model_executor.layers.quantization.utils.w8a8_utils import (

vllm/model_executor/layers/quantization/quark/utils.py

@@ -1,5 +1,6 @@
 import re
-from typing import Any, Optional, Iterable
+from typing import Any, Iterable, Optional
+
 from vllm.model_executor.layers.quantization.utils.quant_utils import (
     FUSED_LAYER_NAME_MAPPING)